JOY PLUS (SILK) DYEING FACTORY LTD. v. CHINESE ARTS & CRAFTS (HK) LTD.

HCA2451/1993

IN THE HIGH COURT OF THE

HONG KONG SPECIAL ADMINISTRATIVE REGION

COURT OF FIRST INSTANCE

HIGH COURT ACTION NO.2451 OF 1993

-------------

-------------

-----------------------

J U D G M E N T

-----------------------

1. The present action arises out of two contracts entered into between the Plaintiff, which carries on the business of a dyeing factory and the Defendant, a manufacturer and seller of garments.

2. The first is Purchase Order No.PO 92-062 dated 3 November 1992 ("the Sales Contract") relating to the supply of 21,300 yards of '85% silk, 15% wool Fuji (fabric washed)' at $75 per yard, half of which was to be delivered on or before 16 November 1992 and the balance on or before 23 November 1992. It was a term of the Sales Contract that colour fastness was to be in accordance with Labtest Hong Kong Limited's Report No.T621079 (TE-28123) dated 20 October 1992. It is common ground that the standard therein specified was :

"COLOUR FASTNESS TO DRY CLEANING (DIN 54024-1971, 30 MINUTES, MECHANICAL WASH AT 30℃, IN PERCHLOROETHYLENE)

COLOUR CHANGE 3-4"

It was also a term of the Sales Contract that the fabric supplied would consist of :

"*25 RED 1,700 YDS

*38 OLIVE GREEN 3,200 YDS

*69 NAVY BLUE 16,400 YDS"

It is common ground that the Sales Contract was varied on 11 November 1992 although there is a dispute as to whether the variation was oral, as the Plaintiff contends, or in writing, as the Defendant contends. In any event, it is common ground that the yardage required was increased by 250 yards for the red, 350 yards for the olive green and 100 yards for the navy blue. As varied, the total purchase price under the Sales Contract was $1,703,865.

3. The second is Purchase Order No.PO 92-065 dated 7 November 1992 ("the Processing Contract"), which contained the following description :

"	DESCRIPTION	QUANTITY
ARTICLE NO. :	14101 FOR NORMAL	21650YDS
	DYEING AND SANDWASH
FINISHING WIDTH :	43-44"
COLOUR FASTNESS :	3-4
COLOUR BREAKDOWN:	.....
ARTICLE NO.:	14101 FOR REACTIVE	3550YDS
	DYEING AND SANDWASH
FINISHING WIDTH :	43-44"
COLOUR FASTNESS :	4-5 (FOR BLACK COLOUR SHOULD BE AS DARK AS POSSIBLE)
COLOUR BREAKDOWN :	.....	"

The fabric in question, silk satin, was to be supplied by the Defendant. The cost for normal dyeing and sandwash was $12 per yard and for reactive dyeing and sandwash was $17 per yard. The Processing Contract was also varied on 11 November. The total price due is $369,696.

4. The Defendant contends that no amount is due because as regards the silk wool fabric, what was delivered was unmerchantable because it did not meet the level of colour fastness agreed. Further, the Defendant alleges that it was a term of the contract that it should have been processed by way of reactive dyeing. Although it is not alleged that the work done under the Processing Contract was in any way defective or unacceptable, the Defendant maintained that the Plaintiff knew that the silk satin fabric was to be used in conjunction with the silk wool fabric so that if the latter was unusable for whatever reason, the former would also be rendered useless. In fact, it emerged from the evidence that the Defendant did receive payment for the garments manufactured from fabric it processed under the Processing Contract. So, belatedly, this line of defence - uselessness - was abandoned but the Defendant nevertheless claims the right to a set-off.

Background

5. The business relationship between the Plaintiff and the Defendant commenced at the latest on 27 March 1992. There are, in evidence, delivery notes dated 27 March, 30 March, 6 April, 8 April and 29 April 1992 from the Plaintiff to the Defendant. The delivery notes are in the Plaintiff's standard form. Printed in Chinese at the bottom left hand corner of each note was the following provision :

"Attention : With respect to the cloth stated in this delivery note, if there is any discrepancy in its quantity, colour or any dyeing defect, the cloth must be returned to us for replacement or re-dyeing within three days from the date of this delivery note. We shall not assume any responsibility once the cloth are cut."

Although counsel for the Defendant had intimated that he would be taking issue with this certified translation in respect of the phrase "dyeing defect", this was not pursued during the trial. Therefore the certified translation must stand.

6. By way of background to the Sales Contract, the Defendant's manager Mr Wilson Leung Wing Yiu gave evidence to the effect that the Defendant received an order from its customer Markwald on 18 August 1992 for garments to be manufactured from a silk wool fabric in three colours ("the silk wool programme"). The fabric was to be made up with 85% silk, 15% wool fuji, 26.55 mm and sandwashed. On 26 August 1992, an order was placed with Carson Barney Limited for the fabric. Although the order was for 16,000 odd yds, Carson Barney was told to produce 1,000 yds as a trial run. The finishing required was described as :

"Dyeing, sandwash, and preshrunk. (Finished fabric shrinkage cannot exceed 1% on warp & welt)."

7. Markwald was not happy with the silk wool fabric produced by Carson Barney : it failed the washing test and the finishing was also not acceptable. A sample was also tested by SGS found it to be below general requirement regarding dimensional stability to washing, dry-cleaning and washing fastness. It is to be noted that in its report of 10 September 1992, SGS had found that on the colour fastness to dry-cleaning (DIN 54024) test, the fabric achieved a level of 3-4. Further samples were received by SGS on 10 September which on testing for colour fastness to dry-cleaning, achieved a level of 4. Nevertheless by 30 September 1992, Markwald had decided to abandon the silk wool programme and, instead, it was persuaded by the Defendant to switch to pure silk as a substitute ("the silk fuji programme"). Markwald agreed subject to testing the finish by dry-cleaning. Problems emerged even on the initial tests on 2 October 1992. After further tests which were also unsatisfactory, Markwald cancelled the silk fuji programme on 7 October 1992.

8. Mr Leung's evidence was that because he knew that Markwald had a preference for the silk wool mixture, he continued to try to find other dyers to perfect the dyeing at a time when Markwald was trying out the substitute 100% silk fuji programme. He said he discussed this matter with Mr Lemke of Markwald who suggested that the fabric should be reactively dyed but that he (Mr Lemke) was not himself an expert. Mr Leung also said that whilst he was in the US on a business trip in September, he enquired about silk wool and in that context, reactive dyeing had been mentioned.

9. On or about 2 October 1992, he called the Plaintiff to ask if they dealt in silk wool fabrics and spoke to Wong Chung Ngai, a shareholder and director of the Plaintiff. Mr Leung said that he told Mr Wong that a colour fastness to washing of a level of 3-4 plus and to dry-cleaning of 4 plus was required. He said he also asked if reactive dyeing could be done to silk wool and that Mr Wong had said that it was possible. A meeting was therefore arranged at the Plaintiff's office for 5 October 1992. Mr Leung went to the meeting accompanied by his assistant Miss Angel Lam. He said he told Mr Wong about Markwald and its requirements, the history of this particular order and what Carson Barney had done. He specified the colour fastness required and said he needed a test to be done for his client who believed in test results. Mr Leung said he specified that reactive dyeing was to be used. Mr Wong's reaction was that this could be done but that it would be more expensive and it would give rise to some problems. Despite this, Mr Leung said that Mr Wong agreed to use the reactive dyeing method. This is denied by Mr Wong and an issue in the case is whether reactive dyeing was a term of the Sales Contract. Mr Leung said that they also discussed setting and shrinkage, that the Plaintiff agreed to set the fabric and proceeded to show the Defendant their setting machine. He said that Choi Yee was mentioned by Mr Wong who said that although Mr Wong only had one setting and shrinking machine, as the delivery dates were tight, in case of need, Choi Yee could be used for shrinkage treatment.

10. When the Defendant received Markwald's cancellation of the silk fuji programme on 7 October 1992, the Defendant sent a fax to Markwald the same day to express his shock and disappointment since the Defendant had already reserved the fabric and had therefore incurred liability. The fax also stated that the Defendant had come up with solutions to some of the problems including colour fastness and that :

"We are now using reactive dye for the dark colours like navy and will do the sandwashing process after dyeing. Reactive dye can achieve very good colour fastness on most of the fabric."

11. There is a further written but undated communication from the Defendant to Markwald after 7 October. It refers to Markwald's decision to cancel the silk wool and its substituted fuji silk order stating that :

"Only yesterday evening we found out the right dyeing factory who knows the nature of silk/wool quality dyeing and they are also holding grey themselves and we requested them to the release as the grey cloth 20,000 metres for your programme."

It also referred to the Defendant having instructed this factory to produce a trial run of 500 metres which could be air-freighted to Markwald for it's perusal by 14 October. A Labtest report would also be given.

12. The Plaintiff produced a test run of navy blue fabric which was delivered on 15 October. A piece of the silk wool fabric as well as a piece of silk satin was submitted to Labtest for testing by the Defendant. The report concluded that :

"BOTH SAMPLES FAIL IN COLOUR FASTNESS TO WASHING. HOWEVER, ALL THE OTHER TEST RESULTS OF BOTH SAMPLES ARE COMMERCIALLY ACCEPTABLE."

It is to be noted that the report is totally silent about reactive dyeing. In fact, there was no reference whatsoever to the method of dyeing. Mr Leung's explanation was that as he had already told Markwald in separate correspondence it was going to be reactive dyeing, and as far as the Plaintiff was concerned, he had already specified reactive dyeing as the method to be used, there was no need to mention it further. Markwald found the sample acceptable and a order was placed as a result. It emerged that Markwald had two skirts made up from the sample it received which were satisfactorily dry-cleaned.

The issues

13. The issues raised in the pleadings are the following :

1. whether it was a term of the Sales Contract that the reactive dyeing method would be used;

2. whether the silk wool fabric met the stipulated level of colour fastness;

3. whether the Defendant has waived any claim it might have or is otherwise estopped from making any claim against the Plaintiff by reason of-

a. the approval of Mr Lo, the quality controller sent by the Defendant to and stationed at the Plaintiff's factory without which the fabrics would not have been delivered;

b. the term in the Purchase Order that all complaints had to be made within three days and in any event before the fabrics were cut, such term being either by custom of the textile trade and/or previous dealings between the parties;

4. whether, if the silk wool fabric delivered under the Sales Contract was unmerchantable, that fact rendered the silk satin fabric delivered under the Processing Contract useless; and

5. whether if the Defendant has a valid defence to the Plaintiff's claim, it has made out its counterclaim in the sum of $1,067,125.08 being loss of profit under a contract with Markwald Gmbh & Co. A.G. ("Markwald"), wasted expenses claimed by Markwald and expenses incurred by the Defendant in relation to that contract with Markwald.

As noted above, Issue (4) above is no longer a live issue.

Reactive Dyeing

14. Was this a term of the Sales Contract? The Sales Contract itself, unlike the Processing Contract, contained no reference to reactive dyeing whatsoever. There is a conflict of evidence regarding what was agreed at the meeting held on 5 October 1992. Whilst Mr Wong accepts that the subject of reactive dyeing was raised and discussed, he said he never agreed to it. Mr Wong said that during that meeting, he pointed out the inherent risks involved in reactive dyeing.

15. Back in 1992, it was not a very popular method of dyeing. Professor Yeung, the Plaintiff's expert has filed a report which supports this statement. From a dyer's point of view, he could expect less difference from tank to tank and roll to roll in normal or acid dyeing as compared to reactive dyeing. The tank to tank difference would be slighter because it is easier to control tank conditions. Reactive dyeing entails more steps and requires a stricter control of conditions.

16. Mr Wong also explained that in 1992, the silk business was good and it was the supplier's market, so they could turn to the easiest method of coming up with the required colour. Mr Wong was insistent that he never agreed to use reactive dyeing, but instead suggested that he produce a first tank as a sample to see if the level of colour fastness could be reached. It was on that basis that a trial run of 300 yards of navy fabric was produced. Mr Wong said the focus was to attain the level of colour fastness that the Defendant's customer required.

17. The Defendant placed considerable reliance on a memorandum dated 11 November 1992 from the Defendant to the Plaintiff which varied both the Sales Contract and the Processing Contract by ordering additional yardage. Against the Sales Contract variation were the words "REACTIVE DYEING". The memorandum refers to an earlier telephone conversation. Miss Lam who gave evidence for the Defendant said that she faxed this memorandum to the Plaintiff and actually confirmed with Miss Yeung with whom she had spoken earlier on the telephone about receipt of the memorandum after she had sent it. Mr Wong said that he never saw the memorandum at the time although he was aware that the Contracts had been varied and that extra yardage had been ordered. He explained the Plaintiff's office system and Miss Yeung's invariable practice to file faxes received. This memorandum could not be located in the Plaintiff's files containing faxes and Miss Yeung had left the Plaintiff's employ some years earlier.

18. It was suggested that Mr Wong was lying because if the Plaintiff had received the memorandum it would be fatal to their case. I disagree. Mr Wong's explanation is entirely plausible and I do not think that the evidence establishes that he was untruthful about not having seen the memorandum at the time. Further, I do not think that the memorandum has the effect the Defendant contends for. Receipt of the memorandum, even if established, is not fatal to the Plaintiff's claim. Either reactive dyeing was a term of the Sales Contract or it was not. It could not have been made a term simply because it appears on the face of the memorandum which was sent on a day when a delivery of about 3,500 yds was made, which was some nine days after the date of the Sales Contract. The memorandum assists the Defendant only in this sense : that it is consistent with their version that it was a term of the Sales Contract.

19. Problems first emerged on about 10 December 1992. Mr Lemke who had been inspecting the finished garments at the Defendant's factory in China found that white gloves he had been wearing during the inspection were badly stained. He therefore took several garments back to his hotel to be dry-cleaned and found that there were problems. But it was not until 11 January 1993 when Markwald forwarded a report from Sandoz to the effect that the fabric was not reactively dyed that it became an issue. The Defendant on receipt of the report sent a sample to Labtest for testing and upon confirmation from Labtest that the fabric had not been reactively dyed, it immediately agreed to compensate Markwald. The Defendant contended that it would not have done so had it not been part of its agreement with Markwald. And if so, it only stands to reason that it was part of the Sales Contract.

20. I do not accept that reactive dyeing was a term in the Sales Contract for the following reasons. First, the Sales Contract made no mention of the requirement of reactive dyeing. In contrast, the Processing Contract specified which part of the order had to be reactively dyed. Second, there is no compelling or convincing reason why Mr Leung should have required the fabric to have been reactively dyed. He himself knew nothing about reactive dyeing, having only heard it mentioned in passing. Whilst Mr Lemke of Markwald had also raised the question of reactive dyeing, he, too, apparently knew little or nothing about that process. There is certainly no evidence that reactive dyeing was the answer to the problems that Markwald had found with the Carson Barney trial run of silk wool fabric, nor of the silk fuji programme which was to be a substitute for the silk wool programme. There is no evidence that the Defendant had ever tried reactive dyeing on the 100% silk fuji notwithstanding their fax to Markwald of 7 October 1992. Third, Labtest was not asked to test the fabric for reactive dyeing. It was the Defendant who arranged for the fabric to be tested and presumably specified the tests that were required. Had it been of importance, it would have featured as one of the tests to be carried out by Labtest. Fourth, reactive dyeing does not feature in the Markwald contract. However, it was accepted that there exists correspondence between the Defendant and Markwald which cannot now be located. This might well explain the arrangement between the Defendant and Markwald. On the evidence, it remains possible that the Defendant agreed to compensate Markwald because of other reasons, such as a business decision to preserve the relationship.

Colour fastness

22. Mr Wong's evidence is that he produced a trial run of the navy silk wool fabric which was delivered to the Defendant in mid-October. The Defendant arranged for this to be tested by Labtest. The fabric was delivered to Labtest on 16 October, and according to their report dated 20 October in so far as it related to the silk wool fabric, it failed the colour fastness to washing test but achieved colour fastness of a level of 3-4 in the DIN 54024 test as set out in the report. A week or so after this report, the Defendant, after consulting its customer Markwald, placed the order but asked the Plaintiff to produce the first tank for the other two colours, namely red and green, which it did by early November.

23. So far as the chronology of events is concerned, deliveries of the fabric were made by the Plaintiff on 11, 16, 19, 21, 25, 27 of November and 1 and 3 December 1992. The larger of the two deliveries made on 21 November 1992 and all subsequent deliveries except one were delivered to Choi Yee on behalf of the Defendant. The delivery position is summarized in the chart below :

* Delivery made to Choi Yee on behalf of the Defendant.

24. At this point I need to digress to deal with the issue of shrinkage treatment. The reason will become apparent.

(a) Shrinkage

25. At the hearing, the Defendant contended that shrinkage was to be undertaken by the Plaintiff under the Sales Contract. The validity of this submission will now be examined. The Sales Contract made no reference to shrinkage as one of the processes to which the fabric to be supplied had to be subjected. By contrast, as noted above, the Carson Barney contract required the fabric to be :

"preshrunk (finished fabric shrinkage cannot exceed 1% on warp & welt)"

26. Although the Labtest report of 20 October 1992 included a test for dimensional stability to washing, it was not incorporated into and made a term of the Sales Contract as was the level of colour fastness. Whilst the evidence of the Defendant was that the order was placed because Markwald and the Defendant found the trial run acceptable, it is not the Defendant's case that this was a sale by sample by reference to the trial run and sale by sample can thus be disregarded.

27. It was part of Mr Wong's evidence that shrinkage affects the cost of production since the price is calculated by reference to the yardage to be supplied. In general, it would mean 5% of the yardage and in terms of the Sales Contract, it would involve just over 1,000 yards or approximately HK$80,000. In the light of this evidence which is unchallenged, had shrinkage treatment been agreed between the parties, it would surely have featured as an express term of the contract.

28. In giving evidence for the Defendant, Mr Leung introduced the concept of shrinkage cum setting. It would appear that he considered 'shrinkage' and 'setting' as interchangeable concepts as well as being inextricably linked, so that one necessarily involved the other. In cross-examination, he eventually accepted that the two processes can be independent and separate. It is plain that neither Mr Leung nor his subordinate Miss Lam, who also gave evidence, really understood the technical aspects involved. They appeared only to be concerned with the fabric being of a consistent width. At the October meeting at the Plaintiff's factory, they had been given a tour of the factory and had been shown the Plaintiff's setting machine which is an expensive and sophisticated piece of computer-operated equipment. Setting is the process to ensure that the yarns are aligned at right angles and is the very last stage or step applied to the fabric since there would be little point in setting the fabric which thereafter had to undergo another process which would upset the alignment. The notion of setting involving also shrinkage may have stemmed from the fact that part of the setting process involved the application of steam to the fabric. But as Mr Wong explained, the quantity and intensity of steam involved in the shrinkage process is far more intense. Moreover, shrinkage can be achieved by home-made machines and there was no need to invest in an expensive piece of machinery that is used for setting running into seven figures. In any event, the amount of steam involved in the setting process is something completely different from what would be involved if the fabric had to be shrunk. In so far as it is suggested that by agreeing to set the fabric the Plaintiff impliedly undertook to shrink the fabric as well, it is to be rejected.

29. As noted above, as from 21 November 1992, most of the deliveries were made to Choi Yee on behalf of the Defendant. The Plaintiff's evidence is that Choi Yee's business is to provide shrinkage treatment for fabrics. The Defendant accepts that the name Choi Yee arose in the initial discussions with Mr Wong in October 1992. He stated that it was made in the context of the tightness of the schedule for delivery and as the Plaintiff only had one setting machine, in case of need, Choi Yee could assist. It is the Defendant's case that deliveries were made to Choi Yee for setting at the request of the Plaintiff because the Plaintiff could not meet its deliveries on schedule. The Plaintiff's case is that it received instructions from the Defendant to make deliveries to Choi Yee for shrinkage treatment. For that reason, fabrics delivered after that date were not set by the Plaintiff as thee would be no point in so doing.

30. There is here a conflict of evidence. On balance, I am inclined to accept the Plaintiff's version. Mr Wong struck me as a witness who knew exactly what he was talking about when it related to know-how. I have no reason to doubt his knowledge of matters within his field. The shrinkage cum setting notion under which Mr Leung and Miss Lam appeared to labour rendered their evidence unreliable when it comes to the technical matters. Had it been the Plaintiff's obligation to preshrink the fabric, the Defendant would not have paid for the services provided by Choi Yee without complaint. It stands to reason that the Defendant would also have claimed for yardage lost through shrinkage treatment. It is to be noted that the Defendant has not produced any invoices from Choi Yee which would throw light on the nature of the services actually provided by Choi Yee. Nor did the Defendant see fit to call anyone from Choi Yee to rebut the Plaintiff's evidence.

31. I find on the evidence that the Plaintiff had no obligation under the Sales Contract to pre-shrink the fabric to be supplied. I also find that deliveries made to Choi Yee were for the purpose of shrinkage treatment. This side issue is of course relevant to the Plaintiff's theory that loss on the grey scale is attributable to shrinkage treatment.

(b) The reports

32. The issue as to whether the fabrics supplied by the Plaintiff attained the contractual level of colour fastness to dry-cleaning specified in Labtest's report of 20 October 1992 has engendered a plethora of reports. On the Plaintiff's side, there are 16 such reports and on the Defendant's side there are 13. Some of the tests were conducted on swatches of fabric, others on finished garments; a certain number were conducted applying an "appearance after dry-cleaning (commercial process)".

33. In addition to the confusing picture thus created, there is also the element of subjectivity inherent in the application of the grey scale. The levels represent the degree of contrast and it is immediately apparent when one looks at the grey scale itself to see how difficult it is to apply the test objectively, particularly in relation to colours other than grey. As I understand it, one has to extrapolate the level of contrast from the grey scale which, as its name indicates, relates only to the colour grey and transpose it (by a process of imagination) to the colour in question.

34. I now turn to consider the reports.

35. Subsequent to the signing of the Sales Contract and up to the period of the last delivery made by the Plaintiff, there are several reports prepared by Labtest. On 3 November 1992, two batches of samples from the first tanks were delivered to Labtest. Mr Wong explained that the first batch was sent simply to obtain the colour fastness to dry-cleaning test. Because the Defendant wanted additional tests viz. on fabric composition, dimensional stability to washing and colour fastness to washing, a further batch of samples was delivered to Labtest later that day.

36. As to the first batch delivered on 3 November 1992, Labtest's report dated 4 November 1992 showed that the red fabric passed the test but not the light olive fabric. It is also apparent from the face of the report that the results of the third sample submitted (navy) were obliterated. Mr Wong explained that this was done because Labtest rang them to say that problems had been discovered concerning the tests, particularly relating to the blue fabric. The test had to be redone and fresh pieces of fabric for all three colours were supplied to Labtest for that purpose. This was why a supplemental report was prepared by Labtest on 17 November 1992 which superceded the 4 November report. The supplemental report shows that all three colours achieved the requisite level of colour fastness. Mr Wong said that he had no recollection of receiving this in 1992 and that it was only after Defendant had raised the query about the accuracy of the original 4 November report that this matter was raised with Labtest who on checking their records found the supplemental report attached to a Report Revision Notification dated 17 November 1992 stating that the 4 November report was superceded by the supplemental report. This was faxed by Labtest to the Plaintiff in July 1995.

37. As to the second batch which required the additional tests, the samples delivered passed the test for colour fastness to dry-cleaning. This report is dated 10 November 1992.

38. On 12 November 1992, the Plaintiff submitted a navy blue sample from the second tank to Labtest which, according to the report dated 17 November, failed the test. Mr Wong's evidence is that this caused the Plaintiff to refix the dye and prepare two more tanks, the third and fourth tanks, which were then submitted for testing on 18 November. Both these tanks passed the test as appears from the report of 19 November. No further tests were conducted by either party until after problems emerged in mid-December 1992.

39. The Plaintiff's submission is that the Plaintiff's reports on tests for colour fastness to dry-cleaning using DIN 54024 all achieved the requisite level. The Defendant sought to challenge the submission in two respects : first, it cast doubt on the accuracy of the 4 November report because the results relating to the blue fabric had been obliterated. But it is obvious from the position of the obliteration in the conclusion section that had it not been obliterated, it would have shown that the navy fabric had also passed the test. It would not have made any sense for the Plaintiff to resort to falsifying a result which would have been in its favour. I am therefore inclined to accept Mr Wong's explanation for the obliteration.

40. Second, as to the 10 November report, the copy which the Defendant received did not show the third column of results for the blue fabric. It was suggested that this rendered the report suspect. The report is identical to the full report save for the missing third column. Mr Wong's evidence is that what was faxed to the Defendant, i.e. the report showing only the results for the red and light olive, was in fact received shortly prior to 10 November as the Plaintiff was anxious to have the results for those two colours from the first tanks in order to proceed with mass production as it had already achieved the requisite level for the blue fabric in the trial run. Whilst I have some difficulty with Labtest providing a 'preview' of a report which is post-dated, the explanation is not entirely implausible, and does not undermine the authenticity of the two versions of the report, the only difference being the absence of third column from one of the versions. In other words, there is no evidence which renders the report(s) of 10 November either false or unreliable.

41. It is to be noted that the report of 17 November relating to the second tank of navy does not detract from the validity of the Plaintiff's submission since it is the Plaintiff's evidence, which was unchallenged, that the dye was refixed for mass production.

42. A number of reports made after the problem had arisen are not helpful in the present analysis because the test applied in those reports was "appearance after dry-cleaning (commercial process)". Prima facie, the test is not that which was stipulated in the Labtest report of 20 October, which is the DIN 54024 test. Further, there is no evidence that the "appearance after dry-cleaning (commercial process)" test is the same as the DIN 54024 test, I therefore do not propose to say more about those reports save that, for reasons not readily apparent, the Defendant failed to supply those reports to the Plaintiff at the time which was during the early stages of the problem emerging.

43. There are two reports from Labtest dated 15 January 1993. One relates to four pieces of fabric (two in blue, one in red and one in green) received from the Defendant on 13 January. The other relates to three pieces of fabric in the three colours received from the Plaintiff on 14 January 1993. The first report shows that the two pieces of navy fabric passed the test but not the green or the red. In the second of the reports, only the red passed the test. It is the Plaintiff's case that the three samples it submitted to Labtest were obtained from the Defendant at a meeting on or about 13 or 14 January after it had been told and sent a copy of the Labtest report of 5 January 1993 showing that all three colours fail the test and that the Defendant retained the other half of each sample to submit for testing. If so, then even when the tests are carried out by the same agency, they may vary by half a grey scale.

44. Mr Leung denied there ever having been such a meeting. He was certainly not present because that apparently coincided with exhibition week and he would have been too occupied with other matters. Mr Leung also rejected any suggestion that at the meeting the samples were cut in half and each party then submitted their sample to Labtest for testing. Leaving aside for the present the question as to whether the samples came from the same garments, the 15 January reports do show that some of the samples tested fail the requisite test. Prima facie therefore, it supports the Defendant's claim that at least some of the fabric used in the manufacture of the garments failed the requisite colour fastness test.

45. There are three other reports submitted by the Defendant to substantiate its case that the fabrics supplied by the Plaintiff did not reach the stipulated level of colour fastness.

46. The first is a report dated 2 February 1993 from Labtest relating to nine garments, all of which were navy in colour. All failed the test : six attained a level of 3 and three attained a level of 2-3.

47. The second is a report made several years later on 2 March 1996 by CMA. The test related to 20 garments randomly drawn from "the goods" by CMA staff. Ten samples were selected on 7 November 1995 and the remaining ten on 10 November 1995. It would appear that the samples were obtained from a warehouse at Yau Tong in Kowloon. In addition to the 20 garments selected, the Defendant supplied seven fabric samples. These were variously described as not having undergone shrinkage and setting finishes, some had been pre-shrunk, others had been reactively dyed or acid-dyed. As the source from which the seven fabric samples were obtained is unclear, occurring several years after the event, and in absence of evidence proving what is asserted in relation to each of the seven samples, I am not prepared to place any reliance on the test relating to the seven samples. Confining myself to the 20 garments and the test results, 13 out of the 20 failed to attain the requisite level : five attained a level of 3, another five attained a level of 2-3, and three attained a level of 2.

48. Finally there is a report of 18 April 1996 by SGS relating to eight pieces of fabric supplied by the Defendant's solicitors. Seven out of the eight attained the requisite level, the only piece that did not was in navy blue which had undergone shrinkage treatment. That attained a level of 3.

49. To counter this evidence, the Plaintiff has put forward a theory to account for the instances of failure. Mr Wong said that he could not understand why two of the three samples he sent for testing on 14 January had failed the test. It occurred to him that shrinkage treatment could possibly have that effect. He had a roll of retained navy fabric cut in half breadthwise and sent one half to Choi Yee for shrinkage. When it came back, he cut a small piece and sent it to Labtest for testing. This resulted in the 12 March report which showed that the fabric failed the test. This was repeated with the retained green fabric. One part was sent to Labtest for testing attaining a level of 4-5. The other part was first sent to Choi Yee for shrinkage, then for testing, this showed a level of 3-4. These reports are dated 25 March and 1 April 1993 respectively.

50. Further, the Plaintiff relies on a report from Ciba-Geigy (Hong Kong) Limited dated 6 April 1993 which states that a comparison of the effect of dry-cleaning on a sample of the navy fabric before and after shrinkage treatment showed a difference of 0.5 on the grey scale. In addition, there are two reports from Professor Yeung, who is the Plaintiff's expert. He conducted certain tests by subjecting pieces of fabric to a steaming process of different periods and concluded that steaming which shrinkage involves, could adversely affect colour fastness to dry-cleaning : the longer the steaming process, the more intense the effect on colour fastness.

51. In my judgment, the Plaintiff has established that shrinkage treatment can have an adverse effect on the level of colour fastness to dry-cleaning. As appears from her letter dated 13 November 1995, even the Defendant's expert, Miss Anita Chan of CMA accepts that :

"additional shrinkage and setting finishes can adversely affect the fastness to a fair extent of about half grade (0.5)."

52. Has the Defendant established its defence on a balance of probabilities?

53. Prima facie, the report of 2 February 1993 and CMA's report of 2 March 1996 substantiate the Defendant's contention. The 18 April 1996 report goes the other way inasmuch as only one out of eight samples failed the test.

54. But there are several respects in which the validity of these reports is undermined. First, there is the adverse effect that shrinkage would have on colour fastness considered above. The difference is at least half a grey scale and possibly more. Second, there is the subjectivity of the test itself and the inherent improbability of the test being applied with any degree of precision. This is demonstrated if not highlighted by the fact that swatches from five of the 20 garments tested by CMA were sent to Springborn for further testing on 22 March 1996. The test results were different in each case : for two of the five, the difference was half a grey scale and for the remaining three, the difference was a full grey scale.

55. Apart from the unreliability of the reports, there are other reasons why the defence has to be viewed with some degree of circumspection. At the time Markwald rejected the goods on 8 January 1993, all the Defendant had were three reports based on the wrong test, namely, "appearance after dry-cleaning (commercial process)". Two of these reports were not even supplied to the Plaintiff. This was then followed by the claim (which I have rejected) that it was a term of the contract that the fabric was to be reactively dyed. By 20 January 1993, not only had the Defendant accepted Markwald's rejection, it had compensated Markwald DM25,000 in respect of expenses incurred by Markwald in relation to the cancelled order. I am troubled by the fact (i) that the initial tests were wrong; (ii) that Markwald's rejection of the goods was based on matters that formed no part of the contract between the Plaintiff and the Defendant; (iii) that such rejection by Markwald was accepted without any protest by the Defendant when there was not then any test that it had commissioned based on DIN 54024 which showed that the blue fabric, which formed 80% of the order, had failed the test.

56. It is also to be noted that garments made up by Markwald from the fabric produced at the trial run which was not reactively dyed was acceptable and attained the level of colour fastness required.

57. Having regard to all these matters, I conclude that the Defendant has failed to establish on a balance of probabilities that the fabric delivered by the Plaintiff did not attain the stipulated level of colour fastness.

Waiver

58. In view of the conclusion I have reached on the first two issues, it is strictly unnecessary to consider the other issues raised. If I were wrong as to those conclusions, these other issues will become relevant. I therefore propose to deal with them briefly.

59. I do not regard the approval of Mr Lo, the quality controller sent by the Defendant to and stationed at the Plaintiff's factory without which the fabrics would not have been delivered as constituting any waiver on the Defendant's part, particularly where the 'defect' in question is not visible to the naked eye and could not readily be detected upon any routine inspection.

The exclusion clause

60. The provision in the delivery notes requiring the goods to be returned within three days if there were any defect or other problems is not part of the contract between the Plaintiff and the Defendant since such a term could not be unilaterally imposed subsequent to the Sales Contract unless it could be said to have been incorporated by a previous course of dealings.

61. The provision may be incorporated by a course of dealing between the parties where each party has led the other reasonably to believe that he intended that their rights and liabilities should be ascertained by reference to the terms of a document which had been consistently used by them in previous transactions. Chitty on Contracts, 27th Ed. Vol.1 at 12-010.

62. In the present case, there are five delivery notes sent to the Defendant in March and April of 1992. There is no evidence of any of the terms of any of the contracts pursuant to which the delivery notes were issued. The earlier transactions, apart from possibly one, involved very limited quantities of fabric. The 'large' order related to only 1615 yards. The work or processes carried out by the Plaintiff consisted of setting in one case, sandwashing in another and sandwashing and setting in a third. The other two do not even contain a description of work done.

63. In these circumstances, I find that the evidence is insufficient to establish that the parties have on previous occasions dealt with each other subject to the provision printed on the delivery notes. For these reasons, I agree with the Defendant that the Plaintiff has failed to establish a course of dealings which would have incorporated that provision as a term of the Sales Contract.

Trade custom

64. The Plaintiff submitted that there was a trade custom that once the fabric has been cut, the customer ceases to have any recourse against the Plaintiff for defective dyeing.

65. There are a number of delivery notes or invoices of other fabric suppliers in evidence that contain a provision similar to that contained in the Plaintiff's delivery note although the period the customer was given to notify the supplier varied from three to seven days. The rationale appears to be that if there is a problem concerning the dye, that can be redone by stripping the original dye and applying new dyes or alternatively the supplier can sell the goods to other buyers who happen to be looking for fabric of that colour. But once the fabric is cut, it becomes a total loss to the fabric supplier.

66. The parties accept that the principle of law stated in Chitty (supra) at 13-014 is correct :

"If there is an invariable, certain and general usage or custom of any particular trade or place, the law will imply on the part of one who contracts or employs another to contract for him upon a matter to which such usage or custom has reference a promise for the benefit of the other party in conformity with such usage or custom; provided there is no inconsistency between the usage and the terms of the contract. To be binding, however, the usage must be notorious, certain and reasonable, and not contrary to law; and it must also be something more than a mere trade practice." (emphasis added)

67. If the provision appearing on the delivery note reflects the trade custom, I have difficulty in upholding it because it cannot be reasonable to expect the customer to be able within the space of three days to ascertain defects, certainly not defects that are not visible to the naked eye.

68. If the trade custom were something different from the printed provision, namely without the time limitation, there is no sufficient evidence to establish such a custom since the provisions contained in the invoices from the other suppliers also have a time limit attached to them. Nor does the evidence of Chan Hak Lai take matters further.

69. Accordingly, the Plaintiff has failed to establish any trade custom that the customer has no recourse against the supplier once the fabric supplied has been cut.

Order

70. For the reasons set out above, the Plaintiff has established its claim against he Defendant. Accordingly, the Defendant is to pay the Plaintiff $1,703,865 in respect of the Sales Contract and $369,696 in respect of the Processing Contract, such sums to carry simple interest pursuant to section 48 of the High Court Ordinance, Cap.4 at 7% per annum from the date of the writ until the date of judgment and thereafter at judgment rate.

71. The Defendant's counterclaim is dismissed.

72. I make an order nisi that the Plaintiff is entitled to the costs of this action and of the counterclaim.